The advent of technologies such as electronic ticketing, social media and mobile phones are providing unprecedented sources of high-volume, highly localised and dynamic flow data in addition to the more typical survey approach. This ‘big data’ phenomenon associated with mobility also brings with it new challenges to the analysis of spatial interactions. Global models of flows resulting from spatial interactions can fail to reveal important local information about their structure, just as global models of other processes in a landscape do. The geographically weighted modelling method has been employed with success in other areas of study to better explore local variations in relationships. The recently described spatially weighted interaction models (SWIM) framework brings a similar focus to local modelling of spatial interactions.